Many decisions in transportation management must be made before the uncertainty of possible traffic conditions is revealed. When making decisions with lasting implications over a medium to long timeframe, it is essential to consider not only the most probable scenario, possibly obtained through a forecasting model but also a range of potential outcomes. This approach allows for effective risk mitigation across a spectrum of scenarios, including less probable ones, and enhances the resilience of planning strategies. In this paper, we demonstrate the development of a generative model capable of learning the multivariate joint probability distribution of link speeds on a road network, using real data collected from sensors. The proposed model has shown its ability to generate scenarios that preserve correlations among variables, while producing samples that faithfully represent the empirical marginal distributions. To further enhance the performance of our Generative Adversarial Network (GAN) model, we employed a Variational AutoEncoder (VAE) for pre-training the generator network. Experimental results, conducted on three distinct benchmark datasets, highlight the potential of the proposed model in generating new scenario samples of multivariate variables. The Wasserstein distance between the generated distribution and the real data, confirms the good performance of our model with respect to state of the art models, based on copulae.

Carbonera, M., Ciavotta, M., Messina, V. (2023). Driving into Uncertainty: An Adversarial Generative Approach for Multivariate Scenario Generation. In Proceedings - 2023 IEEE International Conference on Big Data, BigData 2023 (pp.2578-2587). Institute of Electrical and Electronics Engineers Inc. [10.1109/BigData59044.2023.10386128].

Driving into Uncertainty: An Adversarial Generative Approach for Multivariate Scenario Generation

Carbonera, M;Ciavotta, M;Messina, V
2023

Abstract

Many decisions in transportation management must be made before the uncertainty of possible traffic conditions is revealed. When making decisions with lasting implications over a medium to long timeframe, it is essential to consider not only the most probable scenario, possibly obtained through a forecasting model but also a range of potential outcomes. This approach allows for effective risk mitigation across a spectrum of scenarios, including less probable ones, and enhances the resilience of planning strategies. In this paper, we demonstrate the development of a generative model capable of learning the multivariate joint probability distribution of link speeds on a road network, using real data collected from sensors. The proposed model has shown its ability to generate scenarios that preserve correlations among variables, while producing samples that faithfully represent the empirical marginal distributions. To further enhance the performance of our Generative Adversarial Network (GAN) model, we employed a Variational AutoEncoder (VAE) for pre-training the generator network. Experimental results, conducted on three distinct benchmark datasets, highlight the potential of the proposed model in generating new scenario samples of multivariate variables. The Wasserstein distance between the generated distribution and the real data, confirms the good performance of our model with respect to state of the art models, based on copulae.
slide + paper
Generative adversarial networks; Representation Learning; Road traffic; Uncertainty;
English
IEEE International Conference on Big Data - 15 December 2023 through 18 December 2023
2023
Proceedings - 2023 IEEE International Conference on Big Data, BigData 2023
9798350324457
2023
2578
2587
https://bigdataieee.org/BigData2023/files/BigData2023_ProgramSchedule.pdf
none
Carbonera, M., Ciavotta, M., Messina, V. (2023). Driving into Uncertainty: An Adversarial Generative Approach for Multivariate Scenario Generation. In Proceedings - 2023 IEEE International Conference on Big Data, BigData 2023 (pp.2578-2587). Institute of Electrical and Electronics Engineers Inc. [10.1109/BigData59044.2023.10386128].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/466118
Citazioni
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
Social impact